Tensile properties, tension–tension fatigue and biological response of polyetheretherketone–hydroxyapatite composites for load-bearing orthopedic implants

Biomaterials 24 (2003) 2245–2250

Tensile properties, tension–tension fatigue and biological response of polyetheretherketone–hydroxyapatite composites for load-bearing orthopedic implants M.S. Abu Bakara, M.H.W. Chengb, S.M. Tangc, S.C. Yuc, K. Liaoa, C.T. Tanb, K.A. Khorc,*, P. Cheangc a

Biomedical Engineering Research Center, Nanyang Technological University, Nanyang Avenue, Singapore 639798, Singapore b Department of Orthopaedic Surgery, Singapore General Hospital, Outram Road, Singapore 169608, Singapore c Advanced Materials Research Center, School of Mech. and Prod. Engineering, Nanyang Technological University, Nanyang Avenue, Singapore 639798, Singapore Received 29 October 2002; accepted 9 December 2002

Abstract Polyetheretherketone–hydroxyapatite composites were developed as alternative materials for load-bearing orthopedic applications. The amount of hydroxyapatite (HA) incorporated into the polyetheretherketone (PEEK) polymer matrix ranges from 5 to 40 vol% and these materials were successfully fabricated by injection molding. This study presents the mechanical and biological behavior of the composite materials developed. It was found that the amount of HA in the composite influenced the tensile properties. Dynamic behavior under tension–tension fatigue revealed that the fatigue-life of PEEK–HA composites were dependent on the HA content as well as the applied load. The biological responses of PEEK–HA composites carried out in vivo verified the biocompatibility and bioactive nature of the composite materials. r 2003 Elsevier Science Ltd. All rights reserved. Keywords: Hydroxyapatite; Polyetheretherketone; Bioactive; Composites; Orthopedic; Implants; Mechanical properties

1. Introduction The concept of bioactive particulate reinforced polymer composite as bone analogue was introduced in the early 1980s by Bonfield et al. [1]. The bone analogue material developed mimics the composition of natural bone, a natural composite material comprising mainly of collagen as the organic matrix and mineral apatite as the inorganic reinforcement. Hydroxyapatite (HA) reinforced polyethylene (PE) composites was the earliest bioactive particulate reinforced polymer composite developed. Currently, it is commercially known as HAPEXTM and has been used successfully in clinical situation [2,3]. Following the pioneering work of Bonfield et al., many other bioactive HA-polymer composites were developed using various polymer matrices including among others polymethylmethacry*Corresponding author. Tel.: +65-7905526; fax: +65-7911859. E-mail address: [email protected] (K.A. Khor).

late (PMMA), polyethylmethacrylate (PEMA), BisGMA and PolyactiveTM [4], Polysulfone [5], Polyhydroxybutyrate (PHB) [6,7], Starch-Ethylene Vinyl Alcohol Copolymer (SEVA-C) [8] and Poly l-lactide (PLLA) [9,10]. Polyetheretherketone (PEEK) is a rigid semi-crystalline polymer possessing excellent mechanical properties and bone-like stiffness [11]. Being one of the highest performance engineering thermoplastic currently available, it has received extensive applications for structural and load-bearing functions in the aerospace and marine industries [12]. Additionally, it also has exceptionally good chemical and fatigue resistance, high temperature durability and good wear properties. From the processing perspective, PEEK polymer parts can be fabricated easily using conventional plastic processing equipment [13,14]. For biomedical applications, PEEK offers additional benefits including its ability to be repeatedly sterilized and shaped readily by machining and heat contouring [11,15]. In recent years, PEEK polymer has

0142-9612/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0142-9612(03)00028-0

2246

M.S. Abu Bakar et al. / Biomaterials 24 (2003) 2245–2250

gained much attention in the biomedical field, particularly in the area of load-bearing orthopedic applications [15–23]. Most of these workers, however, investigate mainly on the mechanical and biological performance of biologically inert PEEK composite such as fiber reinforced PEEK composites. As such, there are currently little publications available on bioactive HA reinforced PEEK composite. Based on the bone-like stiffness of PEEK polymer, it is anticipated that by incorporating bioactive HA and varying the amount would result in composite with elastic modulus comparable to that of natural bone and at the same time make the PEEK polymer bioactive. This study aims to investigate the potential of PEEK–HA composites as a possible bone analogue substitute suitable for loadbearing functions. The tensile properties, long-term durability and preliminary biological analysis carried out in vivo will be presented.

2. Materials and methods 2.1. Raw materials In this study, spherical HA was produced in-house via the wet method using calcium hydroxide and orthophosphoric acid as the starting materials. The process of spheroidization involves spray-drying (Model LT-8, Ohkawara Kakohki Co. Ltd.), followed by flamespraying (Model FP-73, Miller Thermal Inc). Processing details have been provided elsewhere [24]. The flame spheroidized HA produced are structurally dense with glassy smooth surface. The physical characteristics of the flame-spheroidized HA are: density, 3.158  103 kg/ m3; particle size, 3–100 mm; mean size, 25.68 mm; and specific surface area, 3.1  105 m2/m3. The polymer matrix used was medium viscosity grade PEEK polymer (450G,Victrex plc). The melting point and density of PEEK polymer were measured to be 342 C and 1.291  103 kg/m3, respectively. 2.2. Processing of PEEK–HA composite PEEK–HA composites with up to 40 vol% of HA were prepared via a series of processes; i.e. compounding, granulating and injection molding. Compounding was carried out in an internal mixer (Haake) at temperature of 360 C and mixing speed at 40 rpm. The time needed for compounding was adjusted accordingly (i.e. between 10 and 20 min), depending on the amount of HA incorporated. The compounded materials were then granulated to an average size of 3– 4 mm irregular-shaped granules. Injection molding was carried out using the Battenfeld BA-300/050CD injection-molding machine, fitted with single feed end-gated cavity mold. Mini dumb-bell shaped tensile specimens

Fig. 1. Geometry and dimension of the mini dumb-bell shaped tensile specimen.

with geometrical dimension shown in Fig. 1 were produced using semi-automatic mode. The main operating parameters used include: barrel temperature, 345– 390 C; nozzle temperature, 395 C; mold temperature, 50–90 C; injection pressure, 11–14 MPa; injection speed, 12–15 (scale of 1–15); and cooling time, 20 s. 2.3. Tensile and fatigue testing Tensile testing was performed according to the ASTM D638 procedures at room temperature using an Instron IX material testing system. The testing parameters used were: load-cell, 100 kN; cross-head speed, 1 mm/min; and gauge-length, 50 mm. An extensometer was fitted to the specimen to provide accurate measurement of the materials elastic modulus. Cyclic fatigue testing was carried out on the mini dumb-bell shaped tensile specimens using Instron FastTrack 8000 machine, operating under load-controlled mode and using sinusoidal waveform. Testing was carried out at room temperature under tension–tension fatigue at frequency of 5 Hz and R-value of 0.1 (R is the ratio of minimum to maximum cyclic load). In order to evaluate the mechanical durability of the composite, the specimens used for cyclic testing were subjected to three different levels of maximum stress loading, i.e. 30%, 50% and 75% of the ultimate tensile strength (UTS). For each composition tested, the average of three readings was presented for both tensile and fatigue testing, respectively. 2.4. Preliminary in vivo analysis For in vivo study, it is worth noting that the result presented is an extract of a separate comprehensive study involving 24 implants with various configurations

M.S. Abu Bakar et al. / Biomaterials 24 (2003) 2245–2250 16

As prepared

Young's Modulus (GPa)

14

Residual Fatigue

12 10 8 6 4 2 0 0

3.1. Tensile and fatigue properties Fig. 2 shows the typical load–displacement curve for the PEEK–HA composite developed. Evidently, it can be seen that the PEEK–HA composite materials

Fig. 2. Typical load–displacement curves of PEEK composites with varying amount of HA.

20

30

40

50

Fig. 3. Plot of Young’s modulus versus HA vol%, revealing the reinforcement effects of HA for tensile testing and residual fatigue testing. Note that for residual fatigue testing, the samples are those that survived 1 million cycles of cyclic loading of 30% UTS.

120 As prepared

100

Residual Fatigue 80 60 40 20 0

3. Results and discussion

10

HA volume (%)

Tensile Strength (MPa)

(i.e. varying porosity, pore size distribution and HA%) for implantation up to 24 months. The main aim of this in vivo investigation is, therefore, to evaluate the material’s biological response and tissue in-growth at an early implantation period. 35 kg pigs were used and porous cylindrical implants each with a diameter of 10 mm and height of 8 mm were prepared using PEEK composite containing 20 vol% of HA. Implants with measured porosity of 60% and having pore size ranging from 300 to 600 mm were produced by leaching of particulate technique employing a suitable pore-forming agent. The implants were sterilized in a steam autoclave. Under sterile conditions, the periosteum on the lateral surface of the distal metaphyseal femur was removed and a cylindrical cavity was created with a 10 mm drill. The implants were then inserted into the cavity taking care not to damage or crush the implant surface. At 6 and 16 weeks, the pigs were sacrificed and the distal femurs containing the implants were removed. The implants were located within the bone after slicing the distal femurs into 5 mm slices. The retrieved samples were analyzed in two ways. One group was decalcified, fixed in paraffin, sliced with a microtome and stained with hematoxylin and eosin for transmitted light microscopical analysis. The other implant group was embedded in plastic resin, sliced and polished with 1 mm diamond paste and analyzed by using a JEOL-5410 Scanning Electron Microscope (SEM).

2247

0

10

20 30 HA volume (%)

40

50

Fig. 4. Effect of HA contents on the tensile strength of PEEK–HA composites obtained from tensile testing and residual fatigue testing. Note that for residual fatigue testing, the samples are those that survived 1 million cycles of cyclic loading of 30% UTS.

exhibited both ductile and brittle behavior, depending on the amount of HA incorporated into the PEEK polymer. It can be seen that increasing the amount of HA resulted in the composite losing its ductility as seen from the composite failure occurring in the elastic region (i.e. for composite with 30 and 40 vol% HA), typical of brittle mode failure. Figs. 3 and 4 show the Young’s modulus and tensile strength of PEEK–HA composite, respectively. It can be seen that the tensile properties of PEEK–HA composites were dependent on the HA content. The reinforcement effect of HA is clearly evident, however, at the expense of tensile strength. This resulted from the inherent characteristics of monolithic HA, a synthetic ceramic material that is typically characterized by its high stiffness and brittleness. In Fig. 5, it showed that the main fracture mechanism occurred through de-bonding of HA particles from PEEK matrix, suggesting poor interfacial interaction between PEEK and HA. Hence, the drop in tensile strength with increasing HA content. It is also

M.S. Abu Bakar et al. / Biomaterials 24 (2003) 2245–2250

2248

Normalised Elastic Modulus

1.1 1.0 0.9 0.8

SHA-0

0.7

SHA-5 SHA-10

0.6

SHA-20 SHA-30 SHA-40

0.5 1.E-02

1.E+00

1.E+02

1.E+04

1.E+06

Log (Cycles)

Fig. 6. Plot of elastic modulus versus number of cycles, revealing the modulus degradation of PEEK–HA composites. Fig. 5. Fracture surface of PEEK composite with 20 vol% HA, suggesting good dispersion and distribution of HA in the PEEK matrix. It can also be seen that the main fracture mechanism is through de-bonding of HA from PEEK polymer matrix.

Table 1 Fatigue-life data for PEEK–HA composites that survived 1 million cycle at different stress loading HA volume (%) 0 5 10 20 30 40 a

Stress load level that survive 106 cycles (MPa) 30% UTS

50% UTS

75% UTS

23.49 20.69 19.42 17.58 14.75 13.66

39.15 No data 32.36 29.29 24.57 No data

58.72 MPa No data a 506,502 cycles at 48.53 MPa a 6,285 cycles at 43.94 MPa a 34,898 cycles at 36.86 MPa No data

Denotes the number of cycles the sample fracture.

worthwhile mentioning that there are also sightings, however, to a lesser extent on the fracture occurring through the HA particles. Comparing the Young’s modulus of the natural bone, it was found that the modulus of PEEK composite (i.e. with HA content of 10–40 vol%) lies within the low to mid range of the natural bone modulus which ranges from 3 to 30 GPa [25]. The fatigue-life data of PEEK–HA composites under tension–tension loading is shown in Table 1. At lower cyclical loading of 30% and 50% UTS, no failures were observed after 106 cycles for all the composition tested. At higher cyclic loading of 75% UTS, PEEK polymer did not fail even after 106 cycles (i.e. at cyclic loading of 58.7 MPa). The stress loading at which PEEK composites with 5–40 vol% HA survived 106 cycles ranges from 13.7 to 32.4 MPa. At loading of 75% UTS, it can be seen that PEEK composites containing 10 to 30 vol% HA fractured between 6,285 to 506,502 cycles at loading values in the range of 36.9 to 48.5 MPa. This suggests that the fatigue-life of PEEK–HA composites is

dependent on the stress loading level and the amount of HA in the PEEK matrix. Residual Young’s modulus and tensile strength of samples that survived 1 million cycles (i.e. after subjecting to cyclic loading of 30% UTS) were evaluated and plotted in Figs. 3 and 4, respectively. It was clearly evident that both the Young’s modulus and tensile strength deteriorated after subjecting the samples to cyclic loading of 30% UTS. Fatigue damage can be evaluated by quantitative measurement of the material’s elastic modulus after subjecting it to cyclic loading. Fig. 6 illustrates the elastic modulus degradation of PEEK–HA composites cyclical loaded at 30% UTS. An observation worth mentioning is the distinctive profile of the curves that occurred in two stages. In the first instance, gradual drop in elastic modulus was seen up to approximately 105 cycles, subsequently, drastic drop in elastic modulus was observed. This finding suggests that the main damage mechanism consists of filler-matrix de-bonding and interfacial micro-cracking.

3.2. Preliminary in vivo analysis Figs. 7 and 8 show the SEM and optical micrograph, respectively, of the PEEK composite with 20 vol% HA after implantation. For period of 6 weeks, it can be seen from SEM micrograph shown in Fig. 7(a) that normal bone is abutting on the implant with no signs of bone growing into the pores of the porous implant. However, optical micrograph 8(a) revealed the presence of fibrovascular tissue growing within the pores of the implant. For period of 16 weeks of implantation, it was clearly evident that areas of mature bone were formed within the pores of the implant as shown in Fig. 8(b). Osteoblast can be seen laying down osteoid and within the lamellar bone, the presence of osteocytes were observed. The remaining areas are filled with fibrovascular tissue providing blood supply to the forming

M.S. Abu Bakar et al. / Biomaterials 24 (2003) 2245–2250

Host Bone

2249

Fibrous vascular tissue

Implant Material (a) Implant-Bone Interface

Implant material

Pore (b)

Pore

(A) (A) Implant Material

Pore

Bone Tissue In-growth

Implant-Bone Interface

Implant Material Host Bone

Bone

(B) Fig. 7. Scanning electron micrographs of porous PEEK–HA composite after implantation for 6 weeks (labelled A) and 16 weeks (labelled B).

bone. SEM images in Fig. 7(b) also confirms that mature bone can be found within the pores of the implant. Mature bone tissue can be observed adjacent to the bone–implant interface in close apposition.

4. Conclusions A new bioactive PEEK composite was developed with high amount, up to 40 vol% HA. For sufficiently loaded PEEK composite containing 20–30 vol% HA, the mechanical characteristics observed are Young’s modulus, 5 to 7 GPa, tensile strength 49–59 MPa and fatiguelife of 24.6–32.4 MPa at 106 cycles, and they approach the regime of that of cortical bone. The use of a novel approach in ensuring maximum exposure of HA particulate and porosity in the composites when fashioned into implants and tested in vivo suggest favorable bioactivity and biocompatibility. Histological study of animal model revealed the presence of

(B)

Fibrous vascular tissue

Fig. 8. Optical micrograph of porous PEEK–HA composite at 6 weeks (labelled A) and 16 weeks (labelled B) after implantation.

fibroblast cells promoting vascularization evident during early stages of implantation. Following this, osteoblastic activities were seen in the formation of osteoid and osteocytes within lamellar bone in developing mature bone at longer implantation periods. These findings suggest promising use of such composites for high loadbearing application in medical implant, devices and structural scaffolds.

References [1] Bonfield W, Grynpas MD, Tully AE, Bowman J, Abram J. Hydroxyapatite reinforced polyethylene—a mechanically compatible implant material for bone replacement. Biomaterials 1981;2:185–6. [2] Downs RN, Vardy S, Tanner KE, Bonfield W. Bioceramics 1991;4:239–46. [3] Bonfield W. Bioceramics 1998;11:37–40. [4] Bonfield W. Bioceramics 1996;9:11–3.

2250

M.S. Abu Bakar et al. / Biomaterials 24 (2003) 2245–2250

[5] Wang M, Yue CY, Chua B. Production and evaluation of hydroxyapatite reinforced polysulfone for tissue replacement. J Mater Sci Mater Med 2001;9:821–6. [6] Boeree N, Dove J, Cooper JJ, Knowles J, Hastings GW. Development of a degradable composite for orthopaedic use— mechanical evaluation of an hydroxyapatite–polyhydroxybutyrate composite material. Biomaterials 1993;14(10):793–6. [7] Luklinska ZB, Bonfield W. Morphology and ultrastructure of the interface between hydroxyapatite–polyhydroxybutyrate composite implant and bone. J Mater Sci Mater Med 1997;8:379–83. [8] Sousa RA, Mano JF, Reis RL, Cunha AM, Bevis MJ. Mechanical performance of starch-based bioactive composite biomaterials molded with preferred orientation. Polym Eng Sci 2002;42:1032–45. [9] Shikinami Y, Okuno M. Bioresorbable devices made of forged composites of hydroxyapatite (HA) particles and poly l-lactide (PLLA). Part II: practical properties of miniscrews and miniplates. Biomaterials 2001;22(23):3197–211. [10] Verheyen CCPM, deWijn JR, van Blitterswijk CA, deGroot K. Evaluation of hydroxylapatite-poly(l-lactide) composites—mechanical behaviour. J Biomed Mater Res 1992;26:1277–96. [11] Victrex plc. PEEK medical product technical literature. Victrex Technology Center, Lancashire, United Kingdom. [12] Cogswell FN, Leach DC. Thermoplastic structural composites in service. Plast Rubber Compos Process Appl 1992;18:249–54. [13] Lustiger A, Uralil FS, Newaz GM. Processing and structural optimization of PEEK composites. Polym Compos 1990;11: 65–75. [14] Sarasua JR, Remiro PM. The mechanical behaviour of PEEK short fibre composites. J Mater Sci 1995;30:3501–8.

[15] Barton AJ, Sagers RD, Pitt WG. Bacterial adhesion to orthopaedic implant polymers. J Biomed Mater Res 1996;30: 403–10. [16] Evans SL, Gregson PJ. Composite technology in load-bearing orthopedic implants. Biomaterials 1998;9:1329–42. [17] Ramakrishna S, Mayer J, Wintermantel E, Kam Leong W. Biomedical applications of polymer-composite materials: a review. Compos Sci Technol 2001;61:1189–224. [18] vander Sloten J, Labey L, van Audekercke R, vander Perre G. Materials selection and design for orthopaedic implants with improved long-term performance. Biomaterials 1998;19: 1455–9. [19] Jockisch KA, Brown SA, Bauer TW, Merrit K. Biological response to chopped carbon-fibre reinforced PEEK. J Biomed Mater Res 1992;26:133–46. [20] Corvelli AA, Biermann PJ, Roberts JC. Design, analysis and fabrication of a composite segmental bone replacement implant. J Adv Mater 1997;2:2–8. [21] Kelsey DJ, Springer GS. Composite implant for bone replacement. J Compos Mater 1997;31:1593–631. [22] Wenz LM, Merritt K, Brown SA, Moet A, Steffee AD. In-vitro biocompatibility of polyetheretherketone and polysulfone composites. Biomed Mater Res 1990;24:207–15. [23] Akay M, Aslan N. Polymeric composite hip-joint prosthesis. Adv Compos Lett 1992;1:74–6. [24] Cheang P, Khor KA. Addressing processing problems associated with plasma spraying of hydroxyapatite coatings. Biomaterials 1996;17:537–44. [25] Bonfield W, Wang M, Tanner KE. Interfaces in analogue biomaterials. Acta Mater 1998;46:2509–18.